Challenges and benefits of employing event history analysis of longitudinal survey data: the case of environmental attitudes
Jul 26, 13:45
Event history analysis of longitudinal data is one of the few methods that allows for testing the stability of attitudes over time, a topic that has been at the forefront of many debates within the social sciences. This paper discusses the methodological challenges of employing such analysis, issues that may arise when some key-features of the survey design are changed between measurements.
The analysis covers wave 1 and wave 4 of “Understanding Society – the UK Household Longitudinal Study” (UKHLS). The sample of the first wave of the UKHLS consisted of 40,000 households and it was designed to be representative of the UK population, if appropriate measures are used. The UKHLS uniquely provides the opportunity to link environmental attitudes of a person with environmental behaviours and put this in household context. However, some changes in survey design may impact the research results and need to be taken into consideration: changes in survey mode across waves (from paper based in wave 1 to computer assisted in wave 4), changes in the structure of the self-completion questionnaire, and changes related to the person who answers the household questionnaire (may not be the same person across waves). This study explores ways of accounting for such changes and screens for intermediate or co-lateral information that may help in controlling for them. This includes paradata and response data not only from wave 1 and 4, when environmental attitudes were measured, but also information from intermediate waves, when the same individuals were interviewed and provided data that can be used to better understand their responses in similar contexts.
The research provides innovative ways of accounting for the complexity of longitudinal surveys, which are not static projects, but keep pace with the new developments of a dynamic world.